The Golgi apparatus is compartmentalized into cisternae that associate with one another generating stacks. The multiple stacks in mammalian cells move to a peri-centrosomal position where cisternae in neighboring stacks fuse in a homotypic fashion (i.e. cis with cis, etc.) to form a compartmentalized membrane network. The lateral linking of stacks into a ribbon-like network confers optimal processing efficiency and plays a role in cell cycle progression and wound healing. A major goal is to determine the basis for specificity and regulation in the linking reaction. This will provide mechanistic understanding of how membrane networks are formed, compartmentalized, and inherited during cell division. It also promises insight into significant, yet vexing, questions about membrane tethering related to activation of tethering upon membrane binding, promotion of trans interactions, and coordination with SNARE-mediate membrane fusion. Our working model is that GM130 recruits GRASP65 to cis cisternae using a C-terminal PDZ ligand to bind the PDZ2 domain of GRASP65. Then, GRASP65 on adjacent cisternae interact in trans via their PDZ1 grooves, which bind in a homotypic fashion via internal PDZ ligands present in GRASP65. This enhances efficiency and fidelity of bringing the membranes into close proximity for SNARE contacts and membrane fusion. We also hypothesize that a parallel reaction on medial cisternae involves recruitment of GRASP55 by golgin 45 and GRASP55 homotypic interactions. Thus, the specificity of dual PDZ interactions in each GRASP isoform establishes and maintains the compartmentalized membrane network. Finally, evidence suggests that signaling pathways regulate these interactions and we hypothesize that a cascade leads to PLK1-mediated GRASP phosphorylation causing inactivation of the internal tethering ligands to fragment the Golgi ribbon and promote mitotic entry. Many aspects of this model are novel and, if verified, we believe, would be groundbreaking in the field. As a means of providing rigorous and detailed tests we propose to determine the structure of the GRASP domain, which is the conserved part of the two isoforms and contains the two PDZ-like domains. We will then carryout tests of specific hypotheses regarding its structure/function relationship using point mutations and three types of assay: purified protein interactions, in vivo organelle tethering, and live imaging after knockdown and rescue. This work will define the interaction interfaces for tethering and for localizing the GRASP complexes to their respective cisternae, reveal the mechanism and functional role of specificity in tethering distinct Golgi cisternae with distinct GRASP isoforms, and uncover the mechanism of GRASP domain phospho-inhibition.